Object centering!… (v2.4)

[sansnom]

🇬🇧 Hi, attached is a short video concerning the centering of objects: in one of them (the left one), there is a mask which partially obscures another object which overflows. The centering is not correct. I had never noticed this “indelicacy”!…

🇫🇷 Bonjour, ci-joint une petite vidéo concernant le centrage d'objets : dans l'un deux (celui de gauche), il y a un masque qui occulte partiellement un autre objet qui déborde. Le centrage n'est pas correct. Je n'avais jamais remarqué cette “indélicatesse” !…

[loukash]

A nicely made screencast for sure!  
(I'll look into the actual subject later…)

[firstdefence]

You could make a transparent ellipse to encapsulate the eye design.

[sansnom]

🇬🇧 Hi @firstdefence, in fact, what bothers me is that the calculation is done on plots not seen or on plots (like here) which are in a mask. This does not seem very “logical” to me and above all practical in use. But hey, your proposal is entirely valid, although it requires taking back the objects. And then, there remains the possibility of centering the elements by eye. 🙂

🇫🇷 Hi @firstdefence, en fait, ce qui me gène c'est que le calcul se fasse sur des tracés non vus ou sur des tracés (comme ici) qui sont dans un masque. Cela ne me semble pas très “logique” et surtout pratique à l'usage. Mais bon, votre proposition est tout à fait valable, bien qu'elle demande à reprendre les tracés. Et puis, il reste la possibilité de centrer les éléments à l'œil. 🙂

[loukash]

3 hours ago, sansnom said:

The centering is not correct.

The Transform Origin button is your friend!  

[loukash]

7 minutes ago, loukash said:

The Transform Origin button is your friend!  

… together with the Alignment Handles.

[thomaso]

8 hours ago, sansnom said:

there is a mask which partially obscures another object which overflows. The centering is not correct. I had never noticed this “indelicacy”!…

It's odd in particular because of the bounding box for this group: it displays a square only, without an extended size for the nested 'MASK' layer. – In V1 I am unable to reproduce your offset alignment result.

Apart from this issue: What am I missing that I don't understand the need for your 'MASK' layer …

… nor how you made the darkest fill (bottom layer) appear outside the 'Mask' in your example?

[sansnom]

🇬🇧 Hi @thomaso, I realized this “indelicacy” while working on an illustration. For the video, I just created an example that summarizes the “problem”. If you watch the video carefully, you will understand how the two objects are created and their differences.
Otherwise, you have perfectly understood that the bounding boxes of ALL OBJECTS are taken into account for the centering calculation.
Finally, centering is by definition “visual”. We could then legitimately expect that this centering (and all other types of calibration) is calculated “visually” on the objects, otherwise, it no longer has any meaning or interest. And, in order for the centering command to apply effectively, being forced to modify objects also makes no sense.

🇫🇷 Bonjour @thomaso, je me suis rendu compte de cette “indélicatesse” en travaillant sur une illustration. Pour la vidéo, j'ai juste créé un exemple qui synthétise le “problème”. Si vous regardez bien la vidéo, vous comprendrez comment sont créés les deux objets et leurs différences.
Sinon, vous avez parfaitement compris que les boundings box de TOUS LES OBJETS sont pris en compte pour le calcul du centrage.
Pour finir, un centrage est par définition “visuel”. On pourrait alors légitimement s'attendre à ce que ce centrage (et tous les autres types calages) se calcule “visuellement” sur les objets, sinon, cela n'a plus aucun sens ni d'intérêt. Et, afin que la commande de centrage s'applique efficacement, être obliger de modifier les objets n'a aussi aucun sens.

[thomaso]

1 hour ago, sansnom said:

For the video, I just created an example that summarizes the “problem”. If you watch the video carefully, you will understand how the two objects are created and their differences.
Otherwise, you have perfectly understood that the bounding boxes of ALL OBJECTS are taken into account for the centering calculation.
Finally, centering is by definition “visual”. We could then legitimately expect that this centering (and all other types of calibration) is calculated “visually” on the objects, otherwise, it no longer has any meaning or interest.

Sorry, in my previous post I have misread your video! (in particular: the 'MASK' layer's x/y position on the page)

Now, after recreating your layout & layer hierarchy in V1, it appears that V1 creates a the bounding box for a Group + nested "Mask" layer differently than V2:

While your V2 video shows the bounding box of the selected Group layer as a square tight around the masked result (which, imho, appears as a V2 improvement) …

… to me in V1 the bounding box of the selected Group layer is extended by the size of the masked ellipse layer. The layer type "Mask" nested in a "Group" seems to cause this different bounding box, not the offset ellipse layer only:

2 hours ago, sansnom said:

And, in order for the centering command to apply effectively, being forced to modify objects also makes no sense.

I fully agree, the displayed bounding box should be reliable and work with its displayed size for alignment, too. Also the user should not be forced to a different workflow just because of a feature/tool not respecting the displayed bounding box.

Nevertheless and just as a suggestion, this specific visual result may get achieved with a less complex layer hierarchy and, for instance, just 3 "Ellipse" layers, without "Group" or "Mask"), not causing the bounding box issue.